Visa Incorporated does not extend credit. Visa does not, in any meaningful sense, take credit risk. Visa does not hold cardholder balances. Visa operates a global network of card-issuing banks and merchant-acquiring banks and charges a small fee — measured in basis points, not percentages — every time a transaction crosses from one to the other. The network is presently used by approximately four-and-a-half billion cardholders. The business is, in its essential form, a toll booth.

Toll booths are an unfashionable kind of business in 2026. They are also, in our experience, the only kind that the protocol consistently rewards. A toll booth has very low capital requirements relative to revenue (Visa's tangible assets are mostly servers and headquarters; the real moat is on the network side, where the cardholders and merchants both have to be present at once). A toll booth grows roughly with the underlying economic activity it sits on top of — global consumer payments, in this case. A toll booth has, structurally, almost no path to disruption that does not also disrupt the entire payment-card system. Visa has been a public company since 2008. The stock has compounded at approximately fifteen percent a year, gross, over that period, with almost no quarter-to-quarter volatility in its operating metrics.

Toll booths are an unfashionable kind of business in 2026.

The protocol, run against Visa's most recent annual filing, produces an Owner Earnings number of roughly twenty-two billion dollars on revenue of approximately thirty-eight billion. That is an Owner-Earnings margin north of fifty-five percent, which is a number we do not see anywhere else in our universe. The five-year forward analyst consensus on EPS growth lands in the low teens. The terminal growth is held at GDP. The discount rate, for Visa, gets no sector premium — payment networks are not technology businesses in the protocol's taxonomy; they are infrastructure businesses operating with a software cost structure, and the protocol does not over-penalise infrastructure.

The resulting intrinsic value comes out at something in the eight-hundreds per share. Mr. Market is, this morning, asking three-twenty-six. That is a Δ% of considerable magnitude. We will be the first to admit that the protocol may be optimistic on this one — fifty-five percent Owner Earnings margins are not, historically, sustainable indefinitely, and the protocol's growth-fade schedule treats them as if they will be. A more sceptical analyst would haircut the terminal multiplier. We do not haircut. We publish the number the protocol produces and let the subscriber decide.

What we will say is this: Visa is the rare business where the disagreement with Mr. Market today is not whether the business is good. The business is, by any reasonable measure, exceptional. The disagreement is over whether eight-hundred dollars per share is too generous, or whether three-twenty-six is too miserly, or whether the real number sits somewhere between. Reasonable people disagree. The patient owner, holding Visa for the next decade, will likely find the question to have been somewhat academic.

— The editors